Endodontically Treated Tooth

Embed Size (px)

Citation preview

  • 8/13/2019 Endodontically Treated Tooth

    1/7

    Influence of adhesive techniques onfracture resistance of endodonticallytreated premolars with various residualwall thicknesses

    Nicola Scotti, DDS,a RiccardoRota, DDS,b Marco Scansetti, DDS,c

    Davide Salvatore Paolino,ING,d Giorgio Chiandussi, ING,eDamiano Pasqualini, DDS,f and Elio Berutti, MD, DDSg

    Dental School Lingotto, University of Turin, Turin, Italy; Politecnico diTorino, Turin, Italy

    Statement of problem. The choice of restorative method is commonly based on the cavity conguration and the residualnumber of cavity walls. However, the residual wall thickness could be a valuable clinical parameter in the choice of restorationfor endodontically treated teeth.

    Purpose.The fracture resistance of endodontically treated premolars was compared with different wall thicknesses restored

    with direct composite resin with and without cuspal coverage and with and without

    ber post insertion.

    Material and methods.This study included 104 intact human maxillary premolars extracted for periodontal or orthodonticreasons. Standardized mesio-occluso-distal cavities were prepared with different palatal wall thicknesses (1.5, 2, and 2.5 mm)and a buccal wall thickness of 2 mm. Teeth were restored with or without a ber post and with or without cuspal coverage.Specimens were subjected to thermocycling (3000 cycles, 5 to 55C) and embedded in polymerized acrylic resin. Teeth weresubmitted to cyclic fatigue followed by a static fatigue test with a universal testing machine; a compressive force was applied30 degrees to the long axis of the teeth until fracture. The results were statistically analyzed by 3-way ANOVA ( a.05).

    Results. Residual wall thickness (P.004), the type of adhesive restoration (P

  • 8/13/2019 Endodontically Treated Tooth

    2/7

    loss. Of the various post systems avail-

    able, metal orber posts are commonly

    recommended for clinical use.5 The

    introduction of adhesive techniques

    and the use of ber-reinforced posts

    in conjunction with composite resin

    foundation restorations have facilitated

    the preservation of tooth structure.6

    Moreover, if the post and core havethe same elastic modulus as root

    dentin, forces will be distributed along

    the long axis of the post, improving the

    fracture resistance of endodontically

    treated teeth in vitro.7-9 The function of

    a ber post is to improve the retention

    of the denitive restoration and to

    distribute occlusal stressors along the

    remaining tooth structure. Posts do

    not strengthen endodontically treated

    teeth,10,11 especially anterior teeth,12

    although some studies have suggested

    that cast metal posts result in reduced

    stresses in the dentin.13

    However, endodontically treated

    teeth restored with direct techniques,

    with or without posts, may sometimes

    present residual compromised cusps

    that could eventually fracture. In these

    patients, reduction and coverage with a

    complete crown is usually indicated to

    prevent coronal fracture.14,15 However,

    the concept of minimal intervention

    dentistry to maximize preservation of the

    tooth structure is gaining popularity in

    the restoration of endodontically treated

    teeth.13 Tooth structure preservation is

    directly correlated with fracture resis-

    tance,16,17 reducing the occurrence of

    catastrophic failures and enhancing the

    longevity of the restoration. In vivo and

    in vitro studies support the use of

    direct intracuspal adhesive techniques

    for the restoration of endodonticallytreated teeth if a normal occlusion

    without parafunction is present.18,19

    This approach has the advantage of

    reinforcing teeth without occlusal

    coverage.20-22 To date, few prospective

    clinical studies have evaluated the

    different types of direct or indirect ad-

    hesive restorations of endodontically

    treated teeth. Mannocci et al23 found

    ber posts and direct composite resin

    restorations to be more effective thanamalgam in preventing root fractures,

    but not secondary caries, of endodonti-

    cally treated premolars. Another pro-

    spective in vivo study24 showed similar

    3-year survival rates between endodon-

    tically treated premolars restored with

    ber posts and direct composite resin

    restorations and complete coverage with

    metal ceramic crowns.

    The thickness of the residual cavitywall could be a simple but effective

    parameter for clinically evaluating the

    remaining tooth structure and conse-

    quently for the selection of the most

    appropriate type of restoration for

    endodontically treated teeth. However,

    this parameter has not been investi-

    gated extensively. The choice of restor-

    ative method is commonly based on the

    cavity conguration and the residual

    number of cavity walls.25,26 However,

    because the residual cavity wall thick-

    ness represents the quantity of remain-

    ing enamel and dentin and is directly

    correlated with the residual sound tis-

    sue,27 wall thickness evaluation could

    provide more information to the clini-

    cian regarding the fatigue resistance

    of an endodontically treated tooth.

    Therefore, the purpose of this in vitro

    study was to evaluate the effect of

    cavity wall thickness on the fracture

    resistance of endodontically treated

    teeth restored with a variety of adhesive

    techniques. The null hypothesis tested

    was that residual wall thickness does

    not inuence the fracture resistance

    or fracture mode of endodontically

    treated premolars restored with or

    withoutber posts and with or without

    cusp coverage.

    MATERIAL AND METHODS

    After receipt of institutional board

    approval, the study included 104 non-

    carious single-rooted maxillary pre-

    molars with mature apices recently

    extracted for orthodontic or peri-

    odontal reasons. The inclusion criteria

    were an absence of restorations, similar

    crown and root sizes, and no cracks

    under transillumination. Selected teeth

    were stored in 0.5% chloramine T tri-

    hydrate at 4

    C. After cleaning the toothsurfaces, endodontic treatment was

    performed on all specimens except

    those in the control group (intact teeth;

    n8). Specimens were endodontically

    instrumented with Pathles (1-2-3) and

    ProTaper (S1-S2-F1-F2-F3) (Dentsply

    Maillefer) to the working length,

    enlarging the apex to a size 30, 0.09

    taper. The specimens were irrigated

    with 5% NaOCl alternated with 10%EDTA (ethylenediaminetetraacetic acid)

    with a 2-mL syringe and 25-gauge

    needle. Specimens were obturated

    with gutta percha by using a heat

    source (DownPack; Hu-Friedy) and an

    endodontic sealer (Pulp Canal Sealer

    EWT; Kerr Corp). Backlling was per-

    formed with the Obtura III system

    (Analytic Technologies).

    After 24 hours, all specimens were

    assigned to 3 groups (W1.5; W2; W2.5;

    32 teeth each) based on the thickness

    of the residual palatal wall (1.5, 2, and

    2.5 mm) after cavity preparation,

    measured with a caliber at the occlusal

    oor of the cavity. The specimens were

    treated as reported by Scotti et al.22 In

    each tooth, a class II mesio-occluso-

    distal cavity was prepared. In order to

    compensate for different coronal anat-

    omy, all teeth were prepared by hand by

    1 experienced operator. The prepara-

    tion was extended until the gingival

    cavosurface margin was 1-mm coronal

    to the cement-enamel junction. The

    residual thickness of the buccal cusps at

    the height of the contour was 2 0.2

    mm. Within each group, the specimens

    were further randomly assigned to 4

    subgroups (8 teeth each) on the basis

    of the restoration type (Table I). In the

    groups with cusp coverage (overlay),

    both cusps were reduced with a

    completelyat enamel margin of up to2 mm.

    In the ber post groups, the post

    space was prepared with drills from the

    post manufacturer (Dentsply Maillefer)

    to a depth of 7 mm, measured from the

    pulpal chamber oor. All adhesive

    procedures were performed with All-

    Bond 3 (Bisco) according to the man-

    ufacturers instructions. For ber

    post cementation, a dual-polymerizing

    cement (NanoCore Dual; Dentalica)was used. Dual-polymerizing cement

    November 2013 377

    Scotti et al

  • 8/13/2019 Endodontically Treated Tooth

    3/7

    was applied into the canal by using a

    tube with a needle and the appropriate

    plug (KerrHawe SA) and by injecting

    the materials into the post spaces with

    a specic Composite-Gun (KerrHawe

    SA). Size 1 RTD ber posts (RTD) were

    cemented to the full depth in the

    prepared post spaces. After initial

    preparation, photopolymerization wasperformed with an LED-polymerization

    light for 40 seconds (Translux; Heraeus

    Kulzer) at 1200 mW/cm2.

    After xing the matrix band with a

    retainer, A2-shaded nanohybrid com-

    posite resin (Venus Diamond; Heraeus

    Kulzer) was placed by using an oblique

    layering technique. Each layer, 1.5 to

    2 mm thick, was polymerized for 20

    seconds with an LED-polymerization

    lamp at 1200 mW/cm

    2

    (Translux Po-wer Blue; Heraeus Kulzer). The matrix

    band and retainer were then removed,

    and postpolymerization was performed

    on the buccal and lingual aspects of the

    boxes for 40 seconds on each side. All

    the restored specimens were nished,

    polished, and then stored in distilled

    water at 37C for 7 days. All specimens

    were subjected to 3000 thermal cycles

    between 5 and 55

    C for 60 secondseach. All specimens were embedded in

    light-polymerizing acrylic resin to 1 mm

    below the cement-enamel junction,

    with a thin layer of polyvinyl siloxane

    (Aquasil; Dentsply Italia) around the

    root to simulate the periodontal liga-

    ment. Specimens were then exposed to

    cyclic loading (Mini Bionics II; MTS

    Systems) with an inclination angle of 30

    degrees to the long axis of the tooth

    and at a frequency of 8 Hz, startingwith a load of 20 N for 5000 cycles,

    followed by a stage of 50 N for 20 000

    cycles. A 10-mm-diameter metal ball

    was used. The site of loading was the

    central ssure of the occlusal surface of

    the restoration in the direction of the

    buccal cusp. The specimens were then

    submitted to the static fracture resis-

    tance test by using a universal testing

    machine (Instron) with a 2-mm-diam-eter steel sphere crosshead welded to

    a tapered shaft and applied to the

    specimens at a constant speed of

    2 mm/min and at an angle of 30 de-

    grees to the long axis of the tooth. The

    forces necessary to fracture each tooth

    were measured in newtons (N).

    Restorable and nonrestorable frac-

    tures were distinguished by using an

    optical stereomicroscope with agree-

    ment by 2 examiners. Restorable fail-ures included fractures above the

    Table I. Specimen characteristics, grouping, mean fracture resistance values (Newtons), and fracture modes ofvarious groups

    Group

    (n[32)

    Residual Wall

    Thickness

    Subgroup

    (N[8) Characteristics

    Fracture

    Resistance

    (mean SD)

    Nonrestorable

    Fracture (%)

    W1.5 1.5 mm W1.5 A Post (-), cuspcoverage (-)

    260.51 82.69 87.5

    W1.5 B Post (), cuspcoverage (-)

    445.71 103.19 62.5

    W1.5 C Post (-), cuspcoverage ()

    793.11 196.56 87.5

    W1.5 D Post (), cuspcoverage ()

    775.98 173.30 87.5

    W2 2 mm W2 A Post (-), cuspcoverage (-)

    313.77 69.61 75

    W2 B Post (), cuspcoverage (-)

    660.59 149.48 50

    W2 C Post (-), cusp

    coverage ()

    801.16 175.28 87.5

    W2 D Post (), cuspcoverage ()

    868.58 139.76 87.5

    W2.5 2.5 mm W2.5 A Post (-), cuspcoverage (-)

    368.22 130.27 87.5

    W2.5 B Post (), cuspcoverage (-)

    676.06 155.56 50

    W2.5 C Post (-), cuspcoverage ()

    821.18 121.45 75

    W2.5 D Post (), cuspcoverage ()

    871.77 150.74 87.5

    Control(n8)

    - - Intact teeth 1098.64 287.86 20

    78 Volume 110 Issue 5

    The Journal of Prosthetic Dentistry Scotti et al

  • 8/13/2019 Endodontically Treated Tooth

    4/7

    cement-enamel junction, meaning that

    the tooth could be restored even in in-

    dividuals with major coronal tissue loss.

    Nonrestorable failures included fracture

    patterns that extended below the

    cement-enamel junction, in which a

    surgical approach would be required to

    perform an adhesive restoration.

    A statistical analysis of variance(3-way ANOVA, including 2-way and

    3-way interactions, a.05) was per-

    formed to evaluate the inuence of the

    residual wall thickness (factor T at

    levels 1.5, 2, and 2.5 mm), the resto-

    ration (factor R at inlay and onlay

    levels), and the ber post (factor P at 2

    levels) on the fracture resistance of the

    tooth. A c2 test was used to compare

    the failure modes of the specimens.

    Differences were considered statistically

    signicant at values ofa.05.

    RESULTS

    The mean fracture resistance values

    and the modes of failure observed in

    the groups are listed in Table I. The

    normal probability plot showed that

    the residuals followed a normal distri-

    bution. None of the tested restorations

    showed similar fracture resistance

    values to the control group, even

    those with residual walls of 2.5 mm.

    Within the restored specimens, ANOVA

    (Table II) showed that the thickness of

    the residual wall (P.004), the type of

    adhesive restoration (P

  • 8/13/2019 Endodontically Treated Tooth

    5/7

    treated premolars. Regardless of wall

    thickness and post placement, teeth

    with cusp cupping (factor restoration

    at level onlay) showed the greatest

    fracture resistance.

    The c2 test showed statistically sig-

    nicant differences in the patterns

    of fractures between the subgroups

    (c2

    18.133; P

  • 8/13/2019 Endodontically Treated Tooth

    6/7

    undergoing root canal treat-

    ment.10,13,27,39 According to Aquilino

    and Caplan,13 although ceramic or

    composite resin onlays could protect

    against fractures, no reports support

    their use in the restoration of posterior

    teeth. Some in vitro studies concluded

    that endodontically treated teeth

    recovered the lost resistance to fractureafter receiving an indirect restoration

    with cusp protection.22 The tendency of

    onlay composite resin restorations to

    receive higher load might be attributed

    to the dispersion of compressive

    stresses. Similar results were found for

    cusp coverage with amalgam,39 which

    signicantly increased the fracture

    resistance of the teeth compared to

    those restored without cusp coverage.

    In contrast to the study by Soares

    et al,40 cuspal reduction restorations

    reduced fracture resistance in mandib-

    ular premolars, probably because of the

    large amount of sound tooth tissue

    removed during specimen preparation.

    Other authors have stated that cusp

    coverage does not strengthen premolars

    restored with composite resin onlay32

    or molars with ceramic restorations.41

    This may be due either to the different

    axial direction of the compressive load

    used in these studies, or to the different

    indirect restoration, which involved

    only a buccal cusp. Moreover, in this

    study, the insertion of a ber post

    within the cuspal coverage restoration

    did not signicantly increase fracture

    resistance, probably as a result of the

    adequate dispersion of compressive

    stresses already provided by the com-

    posite resin overlay.21,26,30

    Regarding failure mode, the combi-

    nation of the ber post with an inlayadhesive restoration led to a higher

    incidence of more favorable failure

    types (P2 mm, less

    invasive treatments such as direct in-

    tracuspal composite resin restoration

    supported by ber-post insertion can

    provide sufcient fracture resistance to

    occlusal loads. In contrast, when the

    residual wall thickness is

  • 8/13/2019 Endodontically Treated Tooth

    7/7

    17. Sidoli GE, King PA, Setchell DJ. An in vitro

    evaluation of a carbon ber-based post and

    core system. J Prosthet Dent 1997;78:5-9.

    18. Steele A, Johnson BR. In vitro fracture

    strength of endodontically treated molars.

    J Endod 1999;25:6-8.

    19. Yamada Y, Tsubota Y, Fukushima S. Effect of

    restoration method on fracture resistance of

    endodontically treated maxillary premolars.

    Int J Prosthodont 2004;17:94-8.

    20. Ausiello P, De Gee AJ, Rengo S, Davidson CL.

    Fracture resistance of endodontically-treated

    premolars adhesively restored. Am J Dent

    1997;10:237-41.

    21. Grandini S, Goracci C, Tay FR, Grandini R,

    Ferrari M. Clinical evaluation of the use of

    ber posts and direct resin restorations for

    endodontically treated teeth. Int J Prostho-

    dont 2005;18:399-404.

    22. Scotti N, Scansetti M, Rota R, Pera F,

    Pasqualini D, Berutti E. The effect of the

    post length and cusp coverage on the cycling

    and static load of endodontically treated

    maxillary premolars. Clin Oral Invest

    2011;15:923-9.

    23. Mannocci F, Bertelli E, Sherriff M,Watson TF, Ford TR. Three-year clinical

    comparison of survival of endodontically

    treated teeth restored with either full cast

    coverage or with direct composite restora-

    tion. J Prosthet Dent 2002;88:297-301.

    24. Mannocci F, Qualtrough AJ, Worthington HV,

    Watson TF, Pitt Ford TR. Randomized clinical

    comparison of endodontically treated teeth

    restored with amalgam or with berpostsand

    resin composite: ve-year results. Oper Dent

    2005;30:9-15.

    25. Nam SH, Chang HS, Min KS, Lee Y,

    Cho HW, Bae JM. Effect of the number of

    residual walls on fracture resistances, failure

    patterns, and photoelasticity of simulatedpremolars restored with or without ber-

    reinforced composite posts. J Endod

    2010;36:297-301.

    26. Mangold JT, Kern M. Inuence of

    glass-ber posts on the fracture

    resistance and failure pattern of endodonti-

    cally treated premolars with varying sub-

    stance loss: an in vitro study. J Prosthet Dent

    2011;105:387-93.

    27. Elayouti A, Serry MI, Geis-Gerstorfer J, Lst C.

    Inuence of cusp coverage on the fracture

    resistance of premolars with endodontic ac-

    cess cavities. Int Endod J 2011;44:543-9.

    28. Assif D, Gorl C. Biomechanical consider-

    ations in restoring endodontically treated

    teeth. J Prosthet Dent 1994;71:565-7.

    29. Torbjorner A, Fransson B. A literature review

    on the prosthetic treatment of structurally

    compromised teeth. Int J Prosthodont

    2004;17:369-76.

    30. Heydecke G, Butz F, Hussein A, Strub JR.

    Fracture strength after dynamic loading of

    endodontically treated teeth restored with

    different post-and-core systems. J Prosthet

    Dent 2002;87:438-45.

    31. Mohammadi N, Kahnamoii MA, Yeganeh PK,

    Navimipour EJ. Effect ofber post and cusp

    coverage on fracture resistance of endodon-

    tically treated maxillary premolars directly

    restored with composite resin. J Endod

    2009;35:1428-32.

    32. Bitter K, Noetzel J, Stamm O, Vaudt J, Meyer-

    Lueckel H, Neumann K, Kielbassa AM. Ran-

    domized clinical trial comparing the effects

    of post placement on failure rate of post-endodontic restorations: preliminary results

    of a mean period of 32 months. J Endod

    2009;35:1477-82.

    33. Boschian Pest L, Guidotti S, Pietrabissa R,

    Gagliani M. Stress distribution in a post-

    restored tooth using the three-dimensional

    nite element method. J Oral Rehabil

    2006;33:690-7.

    34. Siso SH,Hrmzl F, TurgutM, AltundasarE,

    SerperA, ErK. Fractureresistanceof thebuccal

    cusps of root lled maxillary premolar teeth

    restored with various techniques. Int Endod J

    2007;40:161-8.

    35. Bitter K, Meyer-Lueckel H, Fotiadis N,

    Blunck U, Neumann K, Kielbassa AM, et al.Inuence of endodontic treatment, post

    insertion, and ceramic restoration on the

    fracture resistance of maxillary premolars. Int

    Endod J 2010;43:469-77.

    36. St-Georges AJ, Sturdevant JR, Swift EJ Jr,

    Thompson JY. Fracture resistance of pre-

    pared teeth restored with bonded inlay

    restorations. J Prosthet Dent 2003;89:

    551-7.

    37. Allara FW Jr, Diefenderfer KE, Molinaro JD.

    Effect of three direct restorative materials on

    molar cuspal fracture resistance. Am J Dent

    2004;17:228-32.

    38. Cheung GSP, Chan TK. Long term survival of

    primary root canal treatment carried out in a

    dental teaching hospital. Int Endod J

    2003;36:117-28.

    39. Mondelli RF, Ishikiriama SK, de Oliveira

    Filho O, Mondelli J. Fracture resistance of

    weakened teeth restored with condensable

    resin with and without cusp coverage. J Appl

    Oral Sci 2009;17:161-5.

    40. Soares CJ, Soares PV, de Freitas Santos-

    Filho PC, Castro CG, Magalhaes D,

    Versluis A. The inuence of cavity design and

    glass ber posts on biomechanical behavior

    of endodontically treated premolars. J Endod

    2008;34:1015-9.

    41. Stappert CF, Guess PC, Chitmongkolsuk S,

    Gerds T, Strub JR. Partial coverage resto-

    ration systems on molars-comparison of

    failure load after exposure to a

    mastication simulator. J Oral Rehabil

    2006;33:698-705.

    42. Newman MP, Yaman P, Dennison J,Rafter M, Billy E. Fracture resistance of

    endodontically treated teeth restored with

    composite posts. J Prosthet Dent 2003;89:

    360-7.

    43. Chuang SF, Yaman P, Herrero A,

    Dennison JB, Chang CH. Inuence of post

    material and length on endodontically

    treated incisors: an in vitro and nite

    element study. J Prosthet Dent 2010;104:

    379-88.

    44. Chang JW, Soo I, Cheung GS. Evaluation of

    ber post-supported restorations under

    simulated occlusal loading. J Prosthet Dent

    2012;108:158-64.

    Corresponding author:

    Dr Nicola Scotti

    Via Nizza 230, 10100 Turin

    ITALY

    E-mail:[email protected]

    Copyright 2013 by the Editorial Council for

    The Journal of Prosthetic Dentistry.

    82 Volume 110 Issue 5

    The Journal of Prosthetic Dentistry Scotti et al

    http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45mailto:[email protected]:[email protected]://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref45http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref44http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref43http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref42http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref41http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref40http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref39http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref38http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref37http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref36http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref35http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref33http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref32http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref31http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref30http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref29http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref28http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref27http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref26http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref25http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref24http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref23http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref22http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref21http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref20http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref19http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref18http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17http://refhub.elsevier.com/S0022-3913(13)00184-4/sref17